Question about U.S. Educational Performance

I often read in right-of-center sources about how despite the  huge increase in real terms in U.S. spending on public schooling since the early 1980s, test scores haven’t gotten any better.

This strikes me as the types of static analysis that free-market types often criticize in other contexts.  For example, real wages (putting aside the issue of whether we’re measuring them correctly) may be “stagnant” since the late 1970s, but this in part reflects the huge growth in America’s immigrant population, who constitute 15% of all American workers, and tend to be less-educated and lower-paid.  Surely almost all of them are better off than folks from their home countries were in the late 1970s, so at least 15 percent or so of Americans are in that sense “better off” relative to the late 70s.  Relatedly, it makes sense to compare the wages of non-immigrant Americans in 2012 to those of non-immigrant Americans of the late 1970s, rather than mixing demographic apples and oranges.

Similarly with education, it strikes me that there are some major confounding factors that  prevent direct apples to apples comparisons between students today and the 1970s.  There is immigration again, especially of Hispanics, who do much more poorly on average in school than do native-born Americans.  Then there is the rise in single motherhood, divorce rates, and other social trends that tend to reduce student performance.  I’m sure there are other factors that would need to be taken into account.

So I wonder, just out of pure curiosity, whether anyone has tried to do an apples to apples comparison American educational achievement over the last thirty years.   E.g., would a child from an upper-middle class white household with two parents present be expected to score higher or lower or about the same in 2012 as in 1982?  How about a child from a poor Spanish-speaking household with two parents present?  Or a child of Korean descent raised by a single mother?  In short, if we take children with the same demographic characteristics do they really do no better in 2012, or has the additional spending actually created improvements that are masked by changes in the overall demographic profile of American schoolchildren?